Well-intentioned efforts to help people with criminal records get jobs, such as prohibiting employers from asking about such histories, don't appear to be easing racial discrimination in hiring and may be backfiring, an economist said Friday at the Minneapolis Fed.

Among the presenters at a daylong conference hosted by the bank's Opportunity and Inclusive Growth Institute, Rutgers University economist Amanda Agan described research at employers before and after they adopted "ban the box" policies that found applicants with distinctly black names were still getting fewer callbacks for interviews.

The conference marked the latest effort by the institute, formed last year at the Federal Reserve Bank of Minneapolis, to dig at the roots of the persistent gap in employment between whites and blacks. Even with unemployment at the lowest level since the late 1990s in the U.S. and Minnesota, the percentage of unemployed black people has remained twice the level of whites and Latinos.

A growing number of public and large private employers in recent years have tried to tackle the problem by removing barriers early in the hiring process that single out people with criminal records, practices that hit black men hardest. The "ban the box" label for such efforts is tied to eliminating questions, or checkoff boxes, on job applications about criminal records.

Agan conducted an experiment by submitting nearly 15,000 online job applications to employers in New York and New Jersey before and after they had wiped out the criminal history box. She found the employers sought interviews with names that were distinctly white more than distinctly black.

Aaron Sojourner, a University of Minnesota economist, said Agan's research documented how powerful racial discrimination is even in the face of efforts to fight it. In his own presentation, Sojourner said he was exploring whether there was a correlation between an increase in the number of felons in a community and an increase in nonemployment. He said preliminary data suggested there was, but that more research was needed.

Other researchers discussed the ways that government policies, such as requiring licenses and certifications for jobs, constrained opportunities, particularly at a time when job vacancies are high and unemployment low.

Economists Darrick Hamilton of the New School and William Darity Jr. of Duke University discussed their research into the wealth gap between white and black Americans, which they said is rooted in slavery and unkept economic promises after the Civil War.

In 2010, they proposed an asset creation tool nicknamed "Baby Bonds" to reduce inequality.

A newborn baby would be given a trust fund inverse to the net worth of his or her family: $500 for the child of affluent parents and $50,000 for the child of poor parents. The child would not be able to access the fund until age 21 and could only use it for a prescribed asset-enhancing activity, such as funding a debt-free education or buying a house.

Initially, the program would disproportionately aid black Americans because they are disproportionately concentrated at the lower end of wealth distribution. "It would not propel us toward a race-blind or colorblind society," Darity said of the Baby Bonds idea. "But it would propel us to a race-fair, or color-fair, society."

While they would provide a financial start for young adults, the bonds would not be enough to dissuade someone from working, and they would help shrink the gap between the races, he said.

"Since wealth provides insurance against declines in income, particularly unexpected declines, the labor market is not the site that we should emphasize to understand economic inequality," Darity said. "The site that we should emphasize is the processes of stratification that led to sharp wealth differentials."